Being relentlessly resourceful is critical when you’re working on problems with unbounded outcomes. It requires staying the course and constantly trying new things, but it doesn’t mean you have to be the hero of every endeavor.
True resourcefulness is seeing the shape of a problem and identifying what it takes to solve it — even if it means not doing it yourself.
In chemistry, a rate-limiting step is the slowest step in a chain that constrains the overall rate of the reaction.
In product, there is also a success bottleneck — a piece of the puzzle that constrains its success. If you focus on executing that piece well, the product has a real shot in the world. If you don’t, it will flop even if the other pieces are crafted to perfection.
Here are some mistakes I made along the way:
Recommendations gone wrong
When we looked to offer more relevant product recommendations, we came up with all sorts of breathtaking, creative designs. We also verified that customers were already looking to find more selection.
Our data scientist devised a new recommendation engine. We launch an AB test, and found that product views increased, but cart adds were flat, as were checkouts. Customers engaged, but they didn’t buy more or faster. In fact, this new section cannibalized clicks from higher-converting sections of the marketplace!
It’s blatantly obvious in hindsight, but the success of recommendations hinges on having… a strong recommendation engine. Amazon isn’t known for its dazzling design or marketing copy, but their recommendations work because they have lots of data and they effectively use it.
Before you dress up a data-centric feature, understand the limitations in how you’re currently using data, and come up with ways to do it better. Your success depends on expanding data signals, or making better use of your existing signals.
Reorganizing a marketplace
When we decided to overhaul how products were categorized in our marketplace, we spent a lot of time understanding the mental models of our buyers and sellers. We rolled up our sleeves and identified the universe of possible tags a product could have, and redesigned the product activation flow.
As we went deeper, we realized that we desperately needed an operations mastermind to orchestrate all of these changes, including retagging and auditing 1M products. We needed a product operations solution to move forward in order to be confident in the speed and quality of our tagging. That became the success bottleneck.
Getting a customer’s book of business
Many of our sellers already had a book of business with them before joining our marketplace, so we built a product to help them manage and refer their customers.
We needed an easy-to-use interface, compelling copy, and the right incentives to show that we were their partners. We saw promising adoption right out of the gate, but soon noticed that the biggest sellers were holding back.
As much as a beautifully composed product put our best face forward, the most valuable sellers didn’t have the time or desire to poke around — they needed to talk to a real person. Trust and real connections became the success bottleneck. We needed sales to help bridge that gap.
As a product manager, it’s on you to identify how to make a product successful. That requires understanding what will bottleneck its success, and how to erase those constraints. Many times, it requires asking better people than yourself to do it.
The hero complex can be particularly hard for product managers to relinquish because many of us enjoy end-to-end ownership. To go far, we need to love the problem more than our specific contribution to the solution.